Radon exposure and cancers other than lung cancer among uranium miners in West Bohemia.

Recent observations have suggested that radon in the ambient air may cause cancers at sites other than the lung, but the evidence is indirect. We have studied site-specific cancer mortality in 4320 uranium miners in West Bohemia who have been followed-up for an average of 25 years, and in whom a four-fold radon-related excess of lung cancer has already been established. For all cancers other than lung cancer the number of deaths observed was slightly greater than that expected from national rates, but the increase was not significant statistically (ratio of observed to expected deaths [O/E] = 1.11, 95% confidence interval [CI] = 0.98-1.24) and mortality did not increase with duration of employment underground or with cumulative exposure to radon. Non-lung cancer mortality was significantly raised among men who started mining work aged under 25 but the increase was not related to cumulative radon exposure. When twenty-eight individual sites and types of cancer were examined, significantly increased risks were found for cancers of the liver (O/E = 1.67) and gallbladder and extrahepatic bile ducts (O/E = 2.26). For liver cancer, mortality did not increase with duration of employment underground or with cumulative radon exposure. For cancer of the gallbladder and extrahepatic bile ducts, mortality did not increase with duration of employment, but increased with cumulative exposure to radon. Mortality from multiple myeloma, although not significantly increased overall (O/E = 1.08), increased with cumulative exposure to radon. Mortality from leukaemia was not increased overall (O/E = 0.91) and was not related to cumulative radon exposure, but did increase with increasing duration of employment in the mines. There is no evidence in these miners that a radon-rich atmosphere increases the risk of any cancer other than lung cancer. Possible exceptions are cancer of the gallbladder and extrahepatic bile ducts and multiple myeloma but further study is needed before it can be concluded that the associations found are causal